| Commit message (Collapse) | Author | Age |
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
From the user, after the second videos. The stage is 3840 by 2160 at the
same type size: four columns (source, disassembly, state, stack), 69
lines, rows and slots with no lens, 43 with a lens as subject, dumps of
32 rows. Code reads at 1440p and up; the caption, the clock and a lens's
drawing are doubled, and the camera at x2 is the old frame. The stack's
rows carry their offset from rsp and a role. A loupe is the telescope as
a floating panel: the memory round what a pointer leads to, lifted over
the context, a pointer inside it opening the next. The current row is a
white wash, a strike a softened vermilion, held a turned corner; sheet 15
sets the alternatives side by side.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
where its register points
From the designer's review of the SQLite scene. Between far stops the
changed registers strike two frames apart in pwndbg's order, and links
and pins set off with memory. A stack window or a dump that fades in
carries no marks of changes the viewer did not see. At rest a moved
register kept a pin at its old place as well, in every scene: gone. The
kernel's answer is filled before the return arrow leaves, and rax strikes
as it lands. Every dump has one reading, by type (a char quoted, a bool
by name); line numbers are as wide as the program's largest; pins' names
keep the side they came to. Names and measures follow the catalogue
(.loose, theme.metric.calls, time.pin, time.continue_wave).
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
| |
--around BYTES of memory each side of what a pointer leads to, for a
viewer that shows the region and not the word; --disasm ROWS above and
below the pc; and every word of the --stack bytes is followed as a
pointer, not the first sixteen.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The SQLite scene needed blocks the catalogue had no name for; each is
placed. A dump is read by its type or by fields a scene gives it, rows
without one folding into a row that counts them; bytes that are no field
are loose; what a stop did not read is unread and what the kernel took is
given. Addresses and line numbers are as wide as the program needs. The
backtrace can be a lens's drawing, and a lens can change in place. A
continue is a verb of its own: registers strike in a wave, far windows
fade through and arrive without marks, frames two stops share slide. The
kernel's answer is filled before it travels. The calls tree has sizes
past ten leaves and folds past thirteen. Mocks 13 and 14 are two stops of
the SQLite recording.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
| |
The debugger records the call each frame returns from, as an instruction
with its line; canim reads it. main is at list.c:41 while reverse()
runs, not at 42, and the rows of a stop deep in SQLite have their lines.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
| |
A write's buffer was recorded as the syscall's payload only: a reader of
the moment's memory found there what an earlier stop had seen of it
(zeros, in SQLite's page). And record --from with --by syscall took the
loader's first syscall for the place and gave up its breakpoint. Both
found by the first animation of a real program.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Three inserts stopped at sqlite3BtreeInsert with their payloads read off
the caller's stack, the commit's backtrace from unixWrite up to main, the
pwrite64 crossing into the kernel, and page 2 as the kernel took it: its
header, its cell pointers and the three rows, read by field. The scene
is built only where SQLite's source is (-Dsqlite=DIR); its recording is
181 moments, 95 KB.
What a real program asked of the blocks: a dump read by the fields a
scene gives it, with the rows between collapsed; addresses of twelve
digits and line numbers of six; the backtrace as a lens's drawing; a
continue's motion (shared frames slide, a far window fades); memory drawn
only where the stop read it. In the kernel lens a register is framed as
carrying an argument only while it holds it at the stop shown, and an
allocator row is read only where an allocation follows the word.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
| |
The camera's visit at x1.5 cleared the context and the lens's view, and
left the projection from the bytes starting at nothing.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
instruction
The first videos were reviewed frame by frame and the catalogue changed
with them. A lens is a drawing over the program column and its view over
the state column, so the stack keeps six slots, its window on the whole
frame. Operands are whole. list and sort are recorded by instruction
from main (record --from), so their registers strike, links reach nodes
and cells, rows that ran are marked and jumps are drawn. A call's node is
a projection of its return address; pins travel and keep their side; a
mark covers the bytes that changed, at the element's width; a syscall's
result is stored after the kernel returns it and every panel is of one
moment; fast passes are veiled; the clock stands still during a step;
captions get their reading time and a scene's first step comes by four
seconds.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
| |
It was the address in the file, which in a program that is not relocated
is the address again: list!402c89 for 402c89. Now list+2c89.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
An independent review of the first videos, frame by frame, is placed in
the catalogue. A lens is a drawing over the program column and its view
over the state column (the dump; the backtrace for calls), which leaves
the stack six slots as subject: the frame whole, from rsp to past rbp.
Operands are whole. A mark covers the bytes that changed. A pin travels
and keeps its side. A call's node is a projection of its return address.
Fast passes are veiled. Captions have numbers a scene can be checked by,
and a scene's first step comes by four seconds. Effects before causes
and panels from different moments are forbidden by name. Mock 12 draws
fib(4) calling fib(2) from the recording.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
| |
By instruction, a dynamic program's loader and C library start are a
million moments nobody watches: the list of anim/tools from main to its
exit is 3 743.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The blocks, their states and the tokens carry the catalogue's names:
atoms in ui, chain/reg/slot in rows, insn with its line column, jump lane
and annotations (read, write, taken or not, a call with its arguments, a
return, a syscall) in disasm, bytes and hexdump with field paper and the
reading column, the three splits and relayout in stage, the ties (thread,
link and pin, projection, echo). A lens is a band over the whole context,
its drawing left and its bytes right. Frames' spans on the stack, each
instruction's line and a callee's parameter count come from the tracer.
A gallery scene draws every block; the test compares six of its pages
with pictures kept in test/gallery, drawn from a recording kept there
too. The scenes are ported; the motion checks report no pop and no
metronome in any of them.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
An instruction's string carries its source line and, for a direct call,
how many parameters the callee takes (DWARF); telescope.frames gives each
frame's extent on the stack; with --alloc the debugger watches each live
allocation and the allocator's header before it, changes only, 256 KiB a
stop at most. C++ names are shown and found as the source has them (the
command links the C++ runtime for its demangler).
SQLite 3.50.4 through every backend, measured (README, bench/sqlite.sh).
It found two faults: a file compiled by its full path was recorded under
a path joined onto the compile directory, and ztrace run noticed a
target's exit up to 10 ms late.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
The README is now the catalogue: atoms, rows, panels, ties, the frame and
the verbs, each block specified once under the name the code uses, with
rules of composition and what is forbidden. From the user's notes on the
context mocks: a lens is a band across the top, its drawing on the left
and the bytes it is drawn from on the right, a hexdump with each field's
bytes marked and read at the row's end; the disassembly is seventeen rows
in address order with a source-line column and an annotation for what
each row read, set or jumped to; a register is pinned to the node it
points at by name; hex is unbroken; and four kinds of tie say which pane
is a view of which (thread, link, projection, echo).
|
| |
|
|
|
|
|
|
|
|
|
| |
debug info
Events from the target's ring pass a tap on their way to the sink: a
variable declared there (the pass's --instr vars, the agent's ztrace_watch)
joins the regions recorded at every stop. With nothing else instrumented
that is names and types from the compiler, values and everything else from
the debugger. The instrumentation runtime is built stripped, so stepping
by line runs through its calls.
|
| |
|
|
|
| |
names(nr) gives fd, buf, count for read and its like: what a viewer of a
syscall labels an argument with, where it had only the argument's kind.
|
| |
|
|
|
| |
A package that depends on ztrace no longer repeats the plugin's recipe or
searches the install step for the shared agent.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
beside it
A package of its own (anim/), on sketchbook's anim module and pardes's
syntax. A scene opens a recording, takes a cursor into it, and declares
blocks that read the moment the cursor is at: the state column (registers
each followed as a pointer and inked by where it points, the stack
telescope, the backtrace), the program column (the whole source, the
disassembly in address order, and the scene's lens: a node-link list over
a hexdump of its heap, array cells, a call tree, a syscall crossing), and
ties between them. Motion is a step told as a sentence (the line fires,
its effects cascade a level at a time, the mark moves on last), timed by
what the step contains; walk, hurry and sprint pace a loop. Five scenes
from real recordings of fib, a linked list, a sort and syscalls.
First drafts, by the agent's own handover: the frame of the design's
second revision, the hexdump lens and the ties; no disassembly
annotations, no gallery, scenes not yet reviewed against the mocks.
|
| |
|
|
|
|
|
|
| |
A function is decoded once from its start, so the instructions before the
pc are known: each instruction's string now says the length of the one
before it, and ztrace show prints eight above and eight below, as pwndbg
does. A call, a jump or a return at the pc is a branch event with its
target (named) and, for a conditional jump, whether the flags take it.
|
| |
|
|
|
|
|
|
|
|
|
| |
At a stop the instruction at the pc is decoded against the registers it
will run with (capstone's operands): a touch event for each memory operand
(address, size, read or write), what is there recorded with it, and what a
write left there read back at the next stop. An instruction-level
recording so has every memory write of the stepped thread, nothing watched.
The frames of the call chain are recorded and their return addresses
named; telescope.backtrace walks them from the recording and ztrace show
prints them. Process.read no longer panics on address 0.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
The user's premise: show what pwndbg shows, hide nothing; the semantic
views stay as a lens on the machine. So every frame holds the registers
(all, each followed as a pointer), the stack telescope and the backtrace
in a column that never changes, and the source, the disassembly and the
scene's lens in the other. A value's ink says where it points (stack,
heap, code, data; rwx underlined), its paper says what happened to it.
JetBrains Mono NL for everything mono; the floor is mono 20, which a
phone does not read: what narration depends on is also in the caption.
Mocks 9 to 11 are drawn from a real recording of targets/list.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
For a viewer to show what pwndbg's context shows, the debugger records it
while the memory is there: each register and the top of the stack followed
as pointers three deep (--telescope), every code address met named, eight
instructions from the pc disassembled by capstone (built from source for
x86, into the command only; pinned as in 03-codenomicom). zt.telescope
reads a chain back with the kind of memory each link points into, and
ztrace show prints the context, coloured on a terminal as pwndbg colours.
The thirty thousand strings of a small run showed that a string map in
each keyframe costs 700 bytes of tree a string: strings are now one table
for the recording, appended a batch and read whole at open. fib(12) by
instruction with everything on: 9 MB, 71 000 moments/s.
|
| |
|
|
|
|
|
|
|
|
| |
motion.md: the animation principles as they apply here, a timing chart at
60 fps for each kind of event, revised motion tokens, and a timestamped
critique of the five scenes as first rendered. motion/motion.py draws a
per-panel timing chart of a video and tiles its frames into contact
sheets; charts/ and sheets/ are what the critique was read from. Measured:
in fib 75 steps start three or more panels on the same frame, 68 changes
have no frame in between, and the run is 68 evenly spaced steps.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
emit compared a head loaded earlier with a tail read later: when other
producers reserved and the consumer consumed in between, the tail was
past it and the subtraction underflowed. A panic in safe builds (the
ring's own thread test aborted twice in 200 runs; a debug agent in a
threaded target would die the same way), a spurious wait otherwise.
Saturating: not full, and the exchange that follows fails and reloads.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
not a handler's
The kernel starts a signal handler with default PKRU (key 0 only) and
default MXCSR and x87 control word, the thread's own saved in the frame.
The SIGSYS handler makes the thread's syscalls and clones its threads
from there, so under --syscalls:
- a pkey_alloc's new rights were undone when the handler returned;
- a thread was born with no right to keyed memory (V8's compiler
threads fault reading the code space: d8 died of SEGV_PKUERR, and the
35 seconds were its core dump, not a spin);
- a buffer in keyed memory was EFAULT to the kernel, and a fault to the
agent reading it for the event;
- a thread started without its maker's float modes (flush-to-zero).
Both handlers now take the thread's PKRU and control words from the frame
first (signal.adopt), and the SIGSYS one writes PKRU back to the frame
after the syscall. Watched memory is read, and hooked code patched, with
every key open, so a key the thread itself cannot use hides nothing.
test-agent: agent_pkey (a keyed page read, given to syscalls, watched,
read by a later thread; skips itself without keys), and agent_sys's
thread checks its MXCSR. README: --alloc and programs with their own
malloc; the core dump.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
| |
musl's printf and Zig programs write through writev: pardes drawing an
image showed nothing of the kitty graphics commands it sends until this.
The README says what Lua and pardes were run with and what they showed.
|
| |
|
|
|
|
|
|
| |
Lua 5.4 through ztrace cc with everything instrumented, on a ten-line
script: 3.2 million moments, and at a keyframe every 256 the keyframes
were half of its 304 MB. At 4096 it is 156 MB and records in 1.7 s
instead of 3.9; a seek to any moment is a millisecond on average, three at
worst, and reading in order 6 µs a moment. The README has the table.
|
| |
|
|
|
|
|
|
|
| |
Debugger.usdt arms the USDT probes of every mapped module as breakpoints
whose hook records a probe event with its arguments and lets the program
run on, raising a probe's semaphore first; record and dbg take
usdt:PROVIDER:NAME (* for any), running the loader through when the probe
is a library's. glibc's own 78 probes are there for the asking:
memory_sbrk_more shows malloc growing the heap.
|
| |
|
|
|
|
| |
--preload and --alloc on ztrace perf: what the agent says and what the
kernel sees are merged by time into one recording (perf.Sorted, as probe
merges BPF and perf). test-together runs a list under both.
|
| |
|
|
|
|
|
|
|
|
|
| |
Symbols come from Symbols (perf.lookup is gone), string ids from
event.stringId, and ptrace.Process is used as it is now. A USDT
semaphore of the program itself is raised by the tracer: the kernel's
counter landed on the wrong word of a small lld-linked target, whose
page of the file is mapped writable twice. test-kernel also runs the
library tests of perf, usdt and bpf.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
| |
Syscall entries and kprobes send the event alone (a kernel function's
registers are not the thread's; a syscall's are its arguments), a
probe's event is not inherited (it failed the target's clone). The
per-syscall tracepoints are told entry from exit on the events
themselves, not on a second set. The VM has two CPUs and a time limit.
README: the kernel's view, with what each mode costs.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Programs built as instruction arrays for one process (no compiler, no
libbpf), writing Events into a BPF ring buffer; a drop counter in a map;
task_struct offsets from the kernel's BTF; a USDT probe's arguments read
in the kernel at the probe. Without CAP_BPF and CAP_PERFMON it says so;
--dry-run builds and lists everything. Tested offline by running the
programs in a small BPF machine, and for real by test-kernel-vm, which
boots this kernel in QEMU where we are root.
ztrace perf: moments are no longer cut per event unless --cut;
--max-size, and a full recording stops the target and says so.
perf.Sorted merges two backends' events by time.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
perf_event_open on a held child: counters, a sampling profile, every
page fault with its registers, maps, threads, context switches, every
syscall (the per-syscall tracepoints, the only ones open to a user),
and the debug registers as watchpoints and breakpoints that do not stop
the target. One time-ordered stream of Events into a Sink; --threads
follows with a ring per CPU. usdt.zig: the notes of <sys/sdt.h> and
their arguments from a Regs and a memory reader, independent of BPF.
test-kernel runs all of it on targets/kernel_*.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
ztrace record and dbg take --ring, --preload and --alloc: the target gets
the ring ztrace run gives it, and the debugger drains it while the target
runs (polling for the stop, so a full ring cannot hold both) and before it
reports a stop. A program compiled by ztrace cc, with the agent preloaded,
stopped at breakpoints, is one recording in order; test-together checks it.
Recording such a program showed the debugger's file names and the pass's
strings both counted from 1: event.stringId (a hash, top bit set) is now
what every producer without a table of its own uses.
|
| |
|
|
|
| |
Without UBSan the compiler at -O1 folds the read and deletes
aligned_alloc with it, and the list check counts one allocation short.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
zig build bench-agent runs each capability against its native loop and
prints what it costs and writes; README has the table. ztrace run looks
at the ring a little longer before sleeping (a target that finds the
tracer asleep pays a futex wake an event). A hooked function's name is
defined as the string of its address, so the recorder's call stack reads
fib. A static link without libc's malloc says which -u to add. The
archive carries no stack probes, so cc links it in debug builds too.
Tests for --wrap (musl, zig cc) and for a recording that fills.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
| |
A full ring made consumer and producer trade a futex wake for every
record: 2.8 M events/s through ztrace run where the recorder alone does
12 M. One comparison in advance.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
what a recording may hold
Syscall User Dispatch now runs its handler on an alternate stack a thread
(a thread's own stack may not hold a signal frame; node makes such
threads), and makes clone, clone3 and vfork itself: the child leaves
through the gate with the target's registers instead of being stepped
into with the trap flag. Paths and written buffers are cut at 256 bytes
(--capture), a watch or snapshot above 1 MiB is refused, and ztrace run
ends a recording that is full (--max-size) with the target killed and the
file closed.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
libztrace-agent.so (preloaded) and libztrace-agent.a (linked in), one
source: map and thread_start at startup; alloc/free from the interposed
allocator; syscall_enter/exit through Syscall User Dispatch; call/ret
from int3 hooks; mem_write from watched regions diffed in-process; the
target-side API (include/ztrace.h, and the ztrace-agent module for Zig).
ztrace run turns each on (--alloc --syscalls --hook --watch), chooses the
moments (--cut), and stops the target at event kinds without ptrace
(--stop-at, --read, -i). test-agent records real targets and checks the
recordings.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
| |
The gettid per event is gone (a thread-local cache; a forked child resets
it), and environ/getenv are public: the agent reads its ZTRACE_* options
the way the ring's fd was read.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
| |
The consumer cached nothing and re-read head for every record, the
producers tail: the two cache lines bounced on each event. From the LLVM
runtime's profile: an event through the ring 65 ns to 29 ns.
|
| |
|
|
| |
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
llvm/pass.cpp is an LLVM pass plugin: lines, stores, loads, calls and
variables, each call into the runtime carrying names and type descriptions,
selected by globs (-ztrace-only, -ztrace-mem), before the optimizer by
default. src/instr/rt.zig is the runtime (libztrace-rt.a), which also
answers clang's own coverage and -finstrument-functions hooks;
src/instr/types.zig the grammar of the type descriptions and its parser.
ztrace cc adds all of it to a compiler command.
test-llvm compiles fib, two sorts, a linked list and a BFS through ztrace
cc and asserts on their recordings; bench-llvm measures what each selection
costs.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
|
| |
Comments only: call's callee is a name's string id when the backend has
names, a3 its type; ret's aux is the size of the value; a variable's aux
its parameter number; a pc may be 0; a mem_write may hold the first bytes
of a large write.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
|
| |
Session.spawn stores Options.stop_mask in the ring before it forks, so a
tracer that steps a target by its events is not racing the target's first
one.
Co-Authored-By: Claude Fable 5.2 <[email protected]>
|
| |
|
|
|
|
|
| |
Lapis page and vellum ink; a role is a form as well as a colour (current,
write, read, pointer, allocated, freed, fault, kernel); panels as acme
windows; Go Mono for the machine and Go for the narrator; motion as whole
frames at 60 fps. design/mock/gen.py redraws the mocks from the tokens.
|